the salience of race versus socioeconomic status among african american voters

theresa kennedy, princeton university (2014)

INTRODUCTION as majority black and middle class; and Orange, New ichael Dawson’s Behind the Mule, published Jersey as majority black and lower class.7 I analyzed in 1994, examines the group dynamics and the results of the 2008 presidential election in each identity of African Americans in politics.1 of the four townships, and found that both majority MDawson gives blacks a collective consciousness rooted black towns supported Obama at rates of over 90 per- in a history of slavery and subsequent economic and cent.8 The majority white towns still supported Obama social subjugation, and further argues that African at high rates, but not as high as in the majority black Americans function as a unit because of their unique towns.9 I was not surprised by these results, as they shared past.2 Dawson uses data from the 1988 National were in line with what Dawson had predicted about Black Election Panel Survey to analyze linked fate— black group politics.10 However, I was unable to find the belief that what happens to others in a person’s more precise data than that found at the precinct level, racial group affects them as individual members of the so my results could not be specified to particular in- racial group—and group consciousness among blacks dividuals.11 in the political sphere, and then examines the effects The ecological inference problem piqued my in- of black group identity on voter choice and political terest in obtaining individual data to apply my find- leanings. 5, 3 ings at the personal level.13 This issue with ecological Dawson’s work is pertinent today in the years inference is that it assumes that individual-level analy- of the first African-American president, and even sis of results can be understood based upon aggregate more so as African Americans become wealthier and or group results.12 The aggregate data from my prior more integrated into white America.4 Integration has study was only useful at the township level. I wanted bearings on the black group: group ties may become to view linked fate and candidate choice on a person- weaker, specifically in the political realm. As African by-person basis. I felt this relationship would present Americans enter into the middle and upper classes, a more accurate and complete picture of each group’s do they still identify with roots in slavery, segregation, voting decisions. Though the majority vote in each and economic suppression? Are these roots aban- township went for Obama, I wanted to explore what doned for newly elevated class ties? Will most present- subsets of the African-American group were likely day African Americans side with the candidate who voting for Obama, and which might be more inclined looks like them, but doesn’t necessarily protect their to vote for the Republican candidate.15 To do this, I personal economic interests? decided to conduct a survey asking African-American Dawson’s work piqued my interest, especially in voters about their adherence to linked fate and politi- an age in which some academics argue that the ideo- cal leanings. In addition, I was interested in testing logical gap between black and white Americans is whether, in an electoral contest where candidate race shrinking.5 I want to expand on Dawson’s work and is set against socioeconomic status, an African-Amer- see if his idea is still applicable today. I have previously ican voter would be more likely to base his or her vote conducted case studies of four townships in the state upon the candidate’s race or class. I proposed an em- of New Jersey that represent cross sections of race and bedded experiment in my survey that would establish class.6 I categorized Carteret, New Jersey as majority a hypothetical election between four different poten- white and lower class; Morristown, New Jersey as ma- tial candidates of varying race and economic-policy jority white and middle class; Willingboro, New Jersey leanings. This would force the respondent to choose the salience of race versus socioeconomic status among african american voters in race and fiscal policy crossed pair-ups whether a Literature Review and Expectations candidate’s race or socioeconomic status would trump Dawson’s Behind the Mule considers race and the other. socioeconomic status among African Americans as These different race and class pairings are tell- factors for political choice. Dawson describes a cer- ing because of the unique class and economic status tain “New Black Politics… characterized by the trans- typically attributed to African Americans as a group. formation of protest politics into electoral politics 13 Because of the history of slavery as well as the so- with high levels of black political unity… an image of cial, economic, and legal barriers placed on African profound political unity that transcends class.”16 This Americans throughout U.S. history, African Ameri- “new” form of black politics defined by Dawson in cans have been dealt a less opportune political status. 1994 places an emphasis on the group consciousness Lack of economic diversity aided in the formation of a that accompanies African Americans in the political black political group, as fiscal interests were generally arena and at the polls. Dawson discusses this black the same.14 However, I argue that the economic inter- political unity in terms of what he calls a “black utility ests of the black group have begun to vary. Increasing heuristic”17: economic diversity has the potential to separate the It is quite clear that, until the mid-1960s, race was group, and could render Dawson’s findings of the late the decisive factor in determining the opportuni- twentieth century less applicable now, twenty years ties and life chances available to virtually all Af- later. rican Americans, regardless of their own or their I also revisited the four townships I studied to family’s social and economic status. Consequently, it was much more effective for African Americans find out the precinct-level results for the 2012 elec- to determine what was good for the racial group tion, and whether these results are in line with my in- than to determine what was good for themselves dividual-level survey results. I conducted case studies individually. It was more efficient for them to use of the four townships for the more recent election in the status of the group, both relative and absolute, order to see if group consciousness has strengthened as a proxy for individual utility.18 or weakened. This may corroborate or contradict my This black utility heuristic historically assisted survey results. Revisiting these townships was neces- African Americans in choosing which ideologies and sary to make insightful comparisons with the newer political values fit them best as a group. Because Af- individual-level results. rican Americans have a unique history with roots in I have discussed Claudine Gay’s work on the ef- slavery, segregation, and economic subjugation, vot- fects of neighborhood status, quality, and segregation ing as a group has meant more political power and on feelings of linked fate, political group conscious- typically has made sense for most black individuals 15 ness, and the black vote. I am interested in looking struggling with similar fiscal, social, and political is- at how living in close proximity to other blacks af- sues. By working as a group, blacks could advance fur- fects the strength of group unity, and if this unity is ther. Therefore, they would be communally furthering lessened by living in more integrated or mostly white the interests of the entire group. 19 neighborhoods. However, I argue that in the years since Dawson’s In conclusion, I find this research to be com- book was published, perhaps a newer black political pelling due to its implications for understanding the functioning has emerged. Where race used to trump black political community in the United States. I be- socioeconomic status, class status is now gaining in- lieve the study will contribute to a further comprehen- creasing importance for African Americans in the po- sion of the factors affecting political integration and litical realm. In 1994, Dawson wrote: the status of blacks today. There may be further im- plications for the study of affirmative action, neigh- Economic polarization among African Americans has indeed been increasing over the past twenty borhood integration, and more recently, the political years. Both the middle class and the group of eco- effects of racially biased police brutality and mass nomically marginalized African Americans have incarceration. I hope that this study will shed light grown… In the future, the new black middle class on the uniqueness of the African-American political may not identify as strongly with the black com- group, both in the past and today. munity, the Democratic Party, or liberal causes… Many would argue that economic polarization within the black community will continue to in- 81 columbia university journal of politics & society

crease throughout the 1990s and will bring in its and whether it exists in different levels of intensity be- wake increasing political polarization.20 tween different socioeconomic class levels. Dawson, I plan to investigate whether economic stratifica- in his later book (2001), Black Visions: The Roots of tion in the black community has strained the historic Contemporary African-American Political Ideologies, political unanimity of African Americans. Dawson writes, “the great majority of blacks continue, my pre- calls on the “black utility heuristic,” the strength and vious and current work shows, to see their fate linked unity of the black network, and individually linked to that of the race, but how that linkage gets interpret- fate as reasons for sustained black political accord.21 ed is based partly on social position but also partly on However, I suggest that in recent year; political effi- ideological orientation.”28 My study questions whether cacy, ties to the black network, and the sense of com- this notion still holds true today. mon identity among blacks have all begun to dissipate. A Pew Research Center report from 2007, Furthermore, residential separation within the black “Blacks See Growing Values Gap Between Poor and community has become a more frequent occurrence. Middle Class,” argues that blacks are now more segre- Taken together, these two factors may suggest that gated because of widening class differences within the black political unity has begun to weaken.22 racial group:29 “African Americans see a widening gulf The black middle class is distinct from the larger between the values of middle class and poor blacks, black community in the United States. Middle-class and nearly four-in-ten say that because of the diver- African Americans have separated themselves both sity within their community, blacks can no longer be from the lower class and the distant upper class. This thought of as a single race.”30 This claim implies that has led to a ‘pulling away’ of the African American members of the black community may perceive them- class structure at both ends as the top becomes a main- selves as less of a cohesive group due to differences in stream bourgeoisie and the bottom is condemned to class ideals. This is especially significant for black po- ‘“ever-widening poverty.”23 The middle class is forced litical unity. If values are divergent among economic either into isolation in the middle or forced to draw classes, this separates the black vote both ideologically closer to the upper or lower class. Dawson writes that and economically. Not only might this affect black the black middle class tends to mirror the black lower economic choices, but political choices as well. In ad- class economically in some ways, as the government dition, the Pew study found that, “blacks and whites largely employs the middle class. On the other hand, concur that there has been a convergence in the values he cites Kilson (1983), who argues that the black held by blacks and whites.”31 If ideologies between the middle class has begun a process of “status deracial- races are becoming similar, this may imply a political ization.”24 Middle-class blacks have joined the ranks of integration as well, or at least a lessening of diversity higher-class workers, and, as such, the characteristic along racial lines. Within the past twenty years, blacks historical racial identity of economic subjugation be- have also begun to less fervently support black leaders comes convoluted with higher-class economic inter- in politics, religion, and the NAACP less fervently.32 In ests.25 Still, Dawson notes, “[t]he contradictory forces the past, these figures have been unifying individuals, working in the black middle class have led to what but more recently, they have arguably lost some power Kilson (1983) has referred to as the ‘insider/outsider’ in the black community. syndrome. He documents that over 80% of this class The Pew study also cites an interesting contrast still feel racial obligations”.26 Though the middle has to Dawson’s work, which may imply that this change separated itself economically from the African Ameri- is generational: cans as a group, overall it still feels a sense of loyalty to A 54% majority of African Americans say that black political unity. blacks who don’t get ahead are mainly responsible Linked fate is tied together with black group con- for their situation, while just three-in-ten say dis- sciousness as well as with the black utility heuristic. crimination is mainly to blame. As recently as the mid 1990s, when Dawson’s book was published, This concept of linked fate causes African Americans black opinion on this question tilted in the oppo- feel their “fate,” or outcome, is impacted by and tied site direction, with a majority of African Ameri- into that of the African American community as a cans arguing that discrimination is the main rea- whole.27 Through my survey and the rest of my study, son for a lack of black progress.35 I seek to find how strong this sense of linked fate is 82 the salience of race versus socioeconomic status among african american voters Views on this feeling of linked fate seemed to unity. I will analyze questions on linked fate, as well change around the time Dawson’s book was published as hypothetical election results, by age range in order in 1994. This could imply that Dawson’s thesis is, in to see whether age makes a difference in strength of present years, not as salient due to the diffusion of the linked fate. black group. But why would this potential switch oc- In Wypijewski’s 2004 article “Black and Bruised,” cur in the mid-1990s? What changed within this time the changing political loyalties of African Americans frame? are further corroborated. An interviewee in Wypi- In Clemetson’s 2003 article, “Younger Blacks tell jewski’s article argued that “‘Many blacks stay home Democrats to take Notice,” the author takes notice for the election because they feel that neither party is of this same decline in strength of linked fate among attending to their needs… [The] divide in America blacks in recent years. Clemetson points to the age ain’t black and white; it’s the haves and the have-nots, bracket of eighteen- to thirty-five-year olds that were and that’s the truth, darling.’”37 Some African Ameri- not alive during the 1960s Civil Rights Movement.34 cans perceive this divide along class lines as being This is significant in illuminating the pertinence, or even stronger than that along racial lines. In addition, lack thereof, of Dawson’s work today. If eighteen- to neither party always satisfies the entirety of African thirty-five-year olds, the younger population of vot- American voters. Still, one of the women interviewed ing-age blacks, make up almost half of the total eli- in the article states, “No one I met in the 18-to-35- gible black voter population—about 40%—then this year-old cohort said, as did James Sulton, at eighty the age range is extremely important to the black vote lion of one of Orangeburgs’ grand families, 'I’d vote for as a whole.35 In 1988, when the Black National Elec- the Devil if he was a Democrat.’38 This level of com- tion Panel survey was taken, we find that a person mitment to the Democratic Party is arguably strongest aged twenty-seven, the median age for the eighteen among the older members of black America, as these to thirty-five cohort, would have been born in 1961. members grew up through the Civil Rights Movement This means that this individual would have grown up and saw what the Republican Party failed to do at that during the Civil Rights Era. Further, his or her parents time. Today, it may seem that “the only thing anyone may have been active or instilled civil rights values in knows for sure about ‘the black vote’ is that it’s not their children during this time of powerful black soli- monolithic.”39 What once used to be a seamless, stable darity. However, a person that is twenty-seven in 2003, voter group may now be more divided among differ- when Clemetson’s article was written, was born in the ent classes, identities, and values. late 1970s. They missed growing up in the Civil Rights In their 2010 article, “Is Racial Linked Fate Movement by a few years. This younger generation Unique? Comparing Race, Ethnicity Class, Gender, may have been further removed from the Movement, and Religion,” Claudine Gay and Jennifer Hochschild and may not have the same strength of black group discuss the weakening of this group feeling of linked fu- unity as those of the previous generation. tures. Their article states, “In seven surveys from 1984 Because the Civil Rights Movement roughly last- through 2008, the proportion of Blacks who perceive ed from 1955 to 1968, any children born after this time racial commonality ranged between 65 and 83percent, frame may not have experienced a sense of the magni- trending downwards in the 2000s compared with ear- tude of the Movement.36 Black National Election Panel lier decades.”40 The fact that linked fate has weakened survey, which surveyed many who had grown up in in the 2000s points again toward the idea that the Civil this time period, reveals strong black unity and linked Rights Movement may have had strong linked-fate ef- fate among African Americans. If the following gen- fects for those born during that time. As a result, in eration was more likely to take the gains made during more recent years, these effects have waned and have the Civil Rights Movement for granted, then it is also resulted in a collectively weaker sense of linked fate. In possible that unity began to decline after this year. Per- addition, the authors write that the strength of linked haps African Americans are more satisfied with where fate may increase and lessen based on social class.41 I they are today. As the time gap from the 1960s strug- agree that group consciousness and shared fate among gle increases and as African Americans perform better African Americans may be lessened because of differ- economically and socially, they may feel that there is ences in economic preferences and demands. This will less need for a new civil rights Movement or strategic be tested in the survey as well.

83 columbia university journal of politics & society

Gay and Hochschild acknowledge that for blacks, From the previous literature on African Ameri- “cultivating or recognizing the fact that one’s own life cans and linked fate, there are many different camps chances are likely to rise and fall as Blacks gain or lose that attempt to describe African-American political political and social standing enables one to use a few behavior and linked fate. Through my experimental strong cues to make sense of the complex American survey, I will attempt to discern which of these claims racial arena. Perhaps ironically, linked fate in this logic are correct and which are unsupported. I hypothesize is as much a matter of self-interest as of group well- that linked fate has decreased to a degree, especially being.”42 The article demonstrates that perhaps linked among younger African Americans, and to some de- fate is decreasing among the African American popu- gree among middle- and upper class African Ameri- lation. The authors also found that lower-class blacks cans. By analyzing vote choice, linked fate ascriptions, do not necessarily show stronger linked fate than and neighborhood quality, I will test levels of linked upper-class blacks.43 The authors conclude with an fate among different groups of the African American interesting inquiry: is linked fate “primarily a heuris- community today. I plan to find whether race or so- tic or primarily an identity?”44 Asking whether black cioeconomic status is more politically salient for Afri- linked fate functions solely as an agent for the black can Americans. community, or whether it also defines such commu- nity brings us closer to understanding how this char- JUNIOR PAPER: REPLICATION OF RESULTS acteristic affects the political arena. The authors claim My junior independent work analyzed whether that “linked fate perceptions are neither consistently race or socioeconomic status is more salient for Af- nor highly politicized,” but I disagree.45 Especially in rican American voters. To approach this question, I the past, linked fate has arguably had a strong impact analyzed the 2008 presidential election results within on African-American voters, even though this impact four townships of New Jersey. The aggregate results could potentially be decreasing. showed very high levels of linked fate among blacks. Claudine Gay, in her article “Putting Race in The results of my junior paper aligned with Dawson’s Context: Identifying the Environmental Determinants 1994 results; however, I believe that looking at indi- of Black Racial Attitudes,” discusses the effects of racial vidual level data may show that social class ties have segregation in neighborhoods on black political unity. begun to trump race after all, at least in some black She argues that segregation might have significant ef- socioeconomic groups. fects on the perpetuation of these feelings of linked fate and group consciousness.46 Because different levels of Results: 2008 Election socioeconomic status typically experience differing In my junior case study, I analyzed four New Jer- degrees of neighborhood quality, neighborhood qual- sey townships that crossed different majorities of race ity may have an effect on how African Americans view and socioeconomic status in order to discern if one themselves in relation to the larger group. For example, was more important for black voters than the other.49 I even middle-class African Americans will, on average, looked at Orange, Carteret, Willingboro, and Morris- live in worse conditions than white people of similar town. The townships were all comparable in popula- income levels: “For African Americans…the inabil- tion size, and each had similar numbers of blacks and ity to secure favorable residential circumstances may whites as a majority, respectively. The socioeconomi- encourage the belief that race still defines and limits cally similar townships were similar in income level the prospects for socioeconomic attainment”.47 On the as well. Morristown and Carteret were majority white other hand, African Americans of high-income and townships, and Willingboro and Orange were major- class status are less likely to be surrounded by poor ity black. I categorized Morristown and Willingboro living conditions. In turn, they may be less inclined as middle class, and Orange and Carteret as lower to think that being black means suffering injustice.48 class. By crossing both race and class, I could compare This could potentially lead to a decline in linked fate African American group voting habits to white voting among high-income blacks. Therefore I have includ- habits, and middle class to lower class group choices.50 ed a section in my survey asking respondents about neighborhood quality, and I analyze these questions by income brackets.

84 the salience of race versus socioeconomic status among african american voters

Township Population % % Median P o v e r t y I chose to extend the results of my junior paper Size Black White Household rate (%) Income ($) by replicating it for the 2012 election. I found the 2012 Morristown 18,411 14.0 62.5 64,279 9.5 general presidential election results for each of the Carteret 22,844 14.85 50.68 58,614 13.0 above four townships and analyzed this data for any Willingboro 31,629 72.7 17.3 66,479 8.6 trends. I then compared it to the 2008 results. The data Orange 30,134 71.83 12.8 40,818 18.1 found for the recent 2012 election was taken from the county clerk’s records for each county of each town- My hypothesis was that, due to black group con- ship at the precinct level. The results were as follows. sciousness, feelings of linked fate, and ascribing to the black utility heuristic, the African American town- Election Summary Tables (2008 and 2012) ships would vote for Obama at similar, very high rates. Morristown The lower class white township, Carteret, would prob- Year Obama Republican Opponent ably vote along socioeconomic lines. The upper class 2008 69.9% 30.1% white township, Morristown, would do so as well, and 2012 66.67% 31.47% as a result, they would split Obama and McCain, re- spectively.51 Carteret For my results, I used precinct-level data taken Year Obama Republican Opponent from an ArcGIS map created by Stanford University 2008 67.2% 32.8% in 2008. The precinct-level results are in the table be- 2012 73.81% 24.64% low. 52

Election Summary Tables (2008) Willingboro Year Obama Republican Opponent Carteret (Lower class, majority white) 2008 90.9% 9.1% Year Obama McCain 2012 91.24% 7.60% 2008 67.2% 32.8%

Orange (Lower class, majority black) Orange Year Obama Republican Opponent Year Obama McCain 2008 95.1% 4.9% 2008 95.1% 4.9% 2012 96.07 2.84%

Willingboro (Middle class, majority black) Comparing the 2008 data to the 2012 data, the Year Obama McCain following trends were apparent. First, in Morristown, 2008 90.9% 9.1% support for Obama and the Republican opponent each year (John McCain in 2008 and Mitt Romney in 2012) remained about the same. In the Carteret elec- Morristown (Middle case, majority white) Year Obama McCain tions, support for Obama increased about 7 percent- 2008 69.9% 30.1% age points, and support for the Republican opponent decreased by about 8 percentage points. In Willing- As expected, Orange was extremely supportive boro, there was not much change; the support for of Obama at over 95 percent. Willingboro was a bit Obama only increased by about half of 1 percentage lower around 90 percent, perhaps due to the middle- point, and the support for the Republican decreased class fiscal values of many voters fervently clashing by about 1.5 percentage points. And in Orange, the with Obama’s liberal economic policies. The results of support for Obama and the Republican stayed about Morristown and Carteret were both a bit surprising. I the same. This data tells us that there was not much assumed that Morristown would vote along racial as change in support for Obama or the Republican op- well as economic lines. However, Morristown turned ponent outside of Carteret. Still, the fact that there was out at majority rates for Obama. Carteret, which I as- not much change in support for Obama in the ma- sumed would turn out for Obama at high rates due to jority black towns of Orange and Willingboro is not liberal economic ties, was less supportive of him than conclusive evidence for either an increase or decrease the wealthier Morristown.53 in linked fate at the individual level. Replication of Results: 2012 Election 85 columbia university journal of politics & society

vey are also taken from Jackson’s 1988 National Black EXPERIMENTAL SURVEY EXPLANATION Election Panel Survey.58 I hypothesize that neighbor- The survey portion of this study provides the hood racial dynamics may have varying influences on individual results required to combat the ecological different classes of African Americans. inference problem.54 There were 940 completed sur- veys in total. The survey was internet-based and was Experimental Design sent out to African Americans of voting age across The experimental portion of the survey helps the United States by the distributor Survey Sampling uncover the main purpose of this study. I wanted to International (SSI). The survey asked simple demo- find out whether Dawson’s work on African Ameri- graphic questions concerning linked fate, neighbor- cans and political linked fate still holds true today. hood quality, and the hypothetical election. Most of The experiment is set up as a hypothetical presidential the survey questions were taken from the 1988 Na- election in which the candidates vary by race and eco- tional Black Election Panel Survey, the same survey nomic ideology. used as Dawson’s dataset for his 1994 book.55 The sur- There are six experimental conditions that arise vey shed light on African-American feelings of linked from crossing race and economic ideology variables. fate, the quality of their neighborhoods, and whom In the experiment, the race manipulation is simple. they would vote for in a hypothetical election. I chose Each condition varies a name indicating the race of the to analyze the questions among brackets of income candidate as black (Kiara Jackson/Imani Williams) or and age to see if different age and income brackets feel white (Katherine Miller/Molly Harris) along with eco- differently about their connection to the black group nomic ideology (liberal/conservative). I chose these or their feelings about their neighborhood. names based on reports of the most common first and last names given to Caucasians and African Americans Demographic Information in the United States.59,60,61 The economic ideology will The survey begins by asking demographic infor- vary between liberals (Imani Williams/Molly Harris) mation of the respondents in order to create a profile and conservatives (Kiara Jackson/Katherine Miller). of each individual answering the survey. The high- A fiscally liberal candidate is described as solving the lighted questions in this section are income, social budget deficit by raising taxes on the wealthy. The con- class, race, zip code, and political party. These ques- servative candidate supports solving the budget defi- tions reveal some of the targeted information I am cit by reducing wasteful spending on big government looking for from individual respondents. programs. The survey asks respondents which of two hypothetical candidates they would more likely vote Questions on Linked Fate for. The names of the candidates and listings of the The linked fate questions of the survey ask pair-ups are shown in the tables below. whether respondents ascribe to the idea of African American group consciousness and common destiny. Candidate Breakdown These questions are taken from Jackson’s 1988 Nation- Name Perceived Ideological Fiscal policy Abbreviation al Black Election Panel Study, used in Dawson’s 1994 of Candidate Race leaning Katherine White Conservative Reduce WC book, which asks whether African Americans have a Miller government sense of racial group identity that applies to political spending unity and consciousness.56 The respondent will an- Imani Black Liberal Increase BL Williams taxes swer most of the linked fate questions by ranking how Kiara Black Conservative Reduce BC much they agree with a certain statement. Jackson government spending Questions on Neighborhood Quality Molly Harris White Liberal Increase WL taxes The neighborhood questions included in the survey are linked to Claudine Gay’s work on neigh- borhood quality and racial composition as it affects group consciousness, specifically among African Americans.57 The questions from this part of the sur- 86 the salience of race versus socioeconomic status among african american voters Candidate Matchups Experimental Results Candidate Imani Williams Katherine Miller I created tables that cross each hypothetical can- Race Black White didate with her three opponents. The tables show how Fiscal leaning Liberal Conservative well each candidate fared in the election compared to

Candidate Imani Williams Kiara Jackson her three opponents and if varying the identity of the Race Black Black opponent affected the respondent’s vote in the election Fiscal leaning Liberal Conservative involving that fixed candidate. The null hypothesis is that the identity of the opponent has no relationship Candidate Kiara Jackson Katherine Miller to the candidate that is held fixed. To test the null hy- Race Black White pothesis, I conducted chi-squared tests on each race Fiscal leaning Conservative Conservative for all the candidate-opponent tables to check for in- Candidate Katherine Miller Molly Harris dependence. The results are as follows. Note also that Race White White each table repeats one of the elections. Fiscal leaning Conservative Liberal By Income SURVEY RESULTS AND ANALYSIS Imani Williams (BL) Elections I chose to analyze the data by separating them I analyzed the elections by low, middle, and into three income brackets: high, middle, and low-in- high-income brackets for each candidate table. For come. These three brackets result in fairly equal sam- the Imani Williams (BL) table, among low-income ple sizes and represent three differing levels of income respondents, Imani Williams won against all three op- to analyze for linked fate and for the hypothetical elec- ponents. The margin was about 20 percent between tion questions. The range of incomes collected within Imani Williams (BL) and Katherine Miller (WC) (as my study was between $0 and $100,000. By breaking expected among low-income voters), and the mar- down the income brackets into thirds, I came up with gin between Imani Williams (BL) and Kiara Jackson a rough definition of the three income brackets for (BC) was smaller, at about 10 percent. There was a sig- my study. Low-income status is defined as earning an nificant difference between Imani Williams (BL) and annual income of $0 to $29,999; middle-income sta- Molly Harris (WL), with Imani Williams winning 64 tus is defined as having a yearly income of $30,000 to percent of the vote, and Molly Harris just winning 36 $59,000; and high-income status is defined as earning percent. an annual salary of $60,000 to $100,000 or more. For the middle-income category of respondents I created age brackets as well for the three desig- in the Imani Williams (BL) elections, Imani Williams nated age ranges. The oldest bracket includes birthdays beat all of her candidates, and the only significant from 1900 to 1962. The middle bracket includes birth margin was between Imani Williams and Kiara Jack- dates from 1963 to 1980, and the last range is those son. This race yielded a significant p-value, as Imani 1981 to 1994. The last and youngest bracket is around Williams won about 66 percent of votes, and Kiara the eighteen-to thirty-five year old range of “young” Jackson (BC) won about 34 percent. For the high- African Americans who were born after the Civil income respondents in the Imani Williams elections, Rights Movement. This group is composed of those Imani Williams again won all of the elections, but not who did not grow up during or in the immediate af- by a high enough margin to be significant in any of the termath of the Civil Rights Movement. These respon- three elections. The results for each income group in dents have been fairly removed temporally from the the Imani Williams elections are shown in the tables Civil Rights Movement. The middle and older ranges below. both include those born during the Movement. These age groups were chosen because they each made up about one third of the total respondents’ ages, making the samples about equal, and because each represents a different relation to the Civil Rights Movement of the 1960s.

87 columbia university journal of politics & society

Support for Imani Williams (Black liberal) Jackson was significant at the 0.01 level. All income levels In the middle-income category, for the Kiara Opponent name Vote for Imani Vote for p-value Jackson (BC) contests, respondents ranged from mak- Williams opponent ing $30,000 to $59,999 annually. In the race between Katherine Miller 59% 41% 0.01771** (WC) Kiara Jackson and Imani Williams (BL), Kiara Jack- Kiara Jackson 59% 41% 0.02146** son took 33.9 percent, while Imani Williams took (BC) 66.1 percent. In the race between Kiara Jackson and Molly Harris 62% 38% 0.001855*** Katherine Miller (WC), Kiara Jackson took 56.0 per- (WL) cent and Katherine Miller took 44.0 percent. And in Low-income the race between Molly Harris (WL) and Kiara Jack- Opponent name Vote for Imani Vote for p-value Williams opponent son, Kiara Jackson took 30.2 percent of the vote, and Katherine Miller 61% 39% 0.1088 Molly Harris took 69.8 percent. There was statistical (WC) significance for the margin between Imani Williams Kiara Jackson 54% 46% 0.5151 (BL) and Kiara Jackson, and well as for the Molly Har- (BC) ris (WL) and Kiara Jackson election. Molly Harris 64% 36% 0.03565** (WL) The results for the middle-income respondents Middle-income for the Kiara Jackson elections show that middle-class Opponent name Vote for Imani Vote for p-value blacks prefer liberal over conservative candidates. Ki- Williams opponent ara Jackson lost to Imani Williams, with Imani Wil- Katherine Miller 57% 43% 0.3363 liams (BL) taking two-thirds of the vote. The results (WC) were not so different between Kiara Jackson (BC) and Kiara Jackson 66% 34% 0.01109** (BC) Katherine Miller (WC), although Kiara Jackson (BC) Molly Harris 61% 39% 0.1025 still took the majority. And in the Molly Harris (WL) (WL) versus Kiara Jackson contest results among middle- High-income class respondents, Kiara Jackson took 30.2 percent Opponent name Vote for Imani Vote for p-value and Molly Harris took 69.8 percent. These numbers Williams opponent were about the same rates that Kiara Jackson and Mol- Katherine Miller 61% 39% 0.1235 (WC) ly Harris took in the low-income bracket. Kiara Jackson 55% 45% 0.5164 For the high-income category, respondents vot- (BC) ed for Kiara Jackson (BC) 44.7 percent over Imani Molly Harris 61% 39% 0.1317 Williams’ (BL) 55.3 percent of the vote. In the same (WL) income range, the race between Kiara Jackson and Note: * p<0.1, ** p<0.05, *** p<0.01 Katherine Miller (WC) came out to be 60 percent for Percent values approximated to nearest whole number. Kiara Jackson and 40.0 percent for Katherine Miller. And for the race between Kiara Jackson and Molly Kiara Jackson (BC) Elections Harris (WL), Kiara Jackson received 43.5 percent of Next, I analyze the Kiara Jackson (BC) elections the vote, and Molly Harris received 56.5 percent. The for each income bracket. For the low-income category, chi-squared test for the high-income table, among all Kiara Jackson took 45.8 percent of the votes, and her three races, was insignificant. The tables for each in- opponent, Imani Williams (BL), took 54.2 percent of come group in the Kiara Jackson elections are found the vote. Low-income respondents who were given the below. Kiara Jackson and Katherine Miller (WC) matchup voted 62.3 percent for Kiara Jackson and 37.7 percent for her opponent. In the low-income category, for the race between Kiara Jackson and Molly Harris (WL), Kiara Jackson received 27.9 percent of the vote, while Molly Harris received 72.1 percent. In addition, the significance for this test showed a p-value of 0.001091, meaning these results were significant at the 0.01 lev- el. Only the election between Molly Harris and Kiara 88 the salience of race versus socioeconomic status among african american voters Support for Kiara Jackson (Black conservative) In the middle-income division, Molly Harris All income levels took 38.9 percent of the vote, and her opponent Imani Opponent name Vote for Vote for p-value Williams took 61.1 percent. In the race between Molly Kiara Jackson opponent Harris and Katherine Miller (WC), Molly Harris took Imani Williams 41% 59% 0.02146** (BL) 59.6 percent of the vote, and Katherine Miller took Katherine Miller 59% 41% 0.017** 40.4 percent. And in the race between Kiara Jackson (WC) (BC) and Molly Harris, Molly Harris took 69.8 per- Molly Harris 33% 67% 0.00001963*** cent, and Kiara Jackson took 30.2 percent. The p-value (WL) for the Molly Harris versus Kiara Jackson election was Low-income 0.0039, rendering the results for the race significant. Opponent name Vote for Vote for p-value Kiara Jackson opponent The other two races were statistically insignificant. In Imani Williams 46% 54% 0.5151 this class division, Molly Harris took around the same (BL) numbers as in the lower income bracket, and Imani Katherine Miller 62% 38% 0.07415* Williams (BL) did the same; however, the votes were (WC) a few percentage points higher for Molly Harris (WL) Molly Harris 28% 72% 0.0005462*** (WL) and a few lower for Imani Williams. In the election be- Middle-income tween Molly Harris and Katherine Miller (WC), Molly Opponent name Vote for Kiara Vote for p-value Harris took only around 60 percent of the votes, com- Jackson opponent pared to her 72 percent in the lower income elections. Imani Williams 34% 66% 0.01109** Katherine Miller, among middle-class respondents, (BL) took around 40 percent of the votes, but for lower Katherine Miller 56% 44% 0.3961 (WC) class voters, she won only a little over 30 percent of the Molly Harris 30% 70% 0.003919*** votes. (WL) For the high-income respondents who voted in High-income elections in which Molly Harris (WL) was a candi- Opponent name Vote for Kiara Vote for p-value date, the race between Imani Williams (BL) and Molly Jackson opponent Harris turned out 38.6 percent for Molly Harris and Imani Williams 45% 55% 0.5164 (BL) 61.4 percent for Imani Williams. In the race between Katherine Miller 60% 40% 0.138 Molly Harris and Katherine Miller (WC), Molly Har- (WC) ris took 58.3 percent of the vote, and Katherine Miller Molly Harris 43% 57% 0.3763 took 41.7 percent. And in the election between Molly (WL) Harris and Kiara Jackson (BC), Molly Harris took 56.5 Note: * p<0.1, ** p<0.05, *** p<0.01 percent of the vote, and Kiara Jackson took 43.5 per- Values approximated to nearest whole number. cent. The p-values for the high-income respondents in the Molly Harris races were not statistically signifi- Molly Harris (WL) Elections cant. These results showed that high-income respon- The next set of tables I chose to analyze among dents preferred Imani Williams (BL) to Molly Harris different income levels were those pertaining to the at about the same rate as the lower two income groups. elections involving Molly Harris (WL) as a candidate. The tables below show the results for the Molly Harris For respondents who were low-income, 36.2 percent elections. voted for Molly Harris in the race between Molly Har- ris and Imani Williams (BL), and 63.8 percent voted for her opponent. In the Molly Harris versus Katherine Miller (WC) contest, Molly Harris took 67.8 percent of the vote, while Katherine Miller took 32.2 percent. And in the third election between Molly Harris and Kiara Jackson (BC), Molly Harris took 72.1 percent, and Kiara Jackson took 27.9 percent of the vote. The results for all of the low-income tests on Molly Harris elections were statistically significant. 89 columbia university journal of politics & society

Support for Molly Harris (White liberal) All income levels All income Opponent name Vote for Vote for p-value Katherine Miller opponent Opponent name Vote for Molly Vote for p-value Harris opponent Imani Williams 41% 59% 0.01771** (BL) Imani Williams 38% 62% 0.00185*** (BL) Kiara Jackson 41% 59% 0.017** (BC) Katherine Miller 62% 38% 0.002663*** (WC) Molly Harris 38% 62% 0.002663*** (WL) Kiara Jackson 67% 33% 0.00001963*** (BC) Low-income Low-income Opponent name Vote for Vote for p-value Katherine Miller opponent Opponent name Vote for Molly Vote for p-value Harris opponent Imani Williams 39% 61% 0.1088 (BL) Imani Williams 36% 64% 0.03565** (BL) Kiara Jackson 38% 62% 0.07415* (BC) Katherine Miller 68% 32% 0.006258*** (WC) Molly Harris 32% 68% 0.00625*** (WL) Kiara Jackson 72% 28% 0.0005462*** (BC) Middle-income Middle-income Opponent name Vote for Molly Vote for p-value Opponent name Vote for Vote for p-value Harris opponent Katherine Miller opponent Imani Williams 39% 61% 0.1025 Imani Williams 43% 57% 0.3363 (BL) (BL) Katherine Miller 60% 40% 0.1655 Kiara Jackson 44% 56% 0.3961 (WC) (BC) Kiara Jackson 70% 30% 0.003919*** Molly Harris 40% 60% 0.1655 (BC) (WL) High-income High-income Opponent name Vote for Vote for p-value Opponent name Vote for Vote for p-value Molly Harris opponent Katherine Miller opponent IImani Williams 39% 61% 0.1317 Imani Williams 39% 61% 0.1235 (BL) (BL) Katherine Miller 58% 42% 0.2482 Kiara Jackson 40% 60% 0.138 (WC) (BC) Kiara Jackson 57% 43% 0.3763 Molly Harris 42% 58% 0.2482 (BC) (WL) Note: * p<0.1, ** p<0.05, *** p<0.01 Note: * p<0.1, ** p<0.05, *** p<0.01 Values approximated to nearest whole number. Values approximated to nearest whole number.

Katherine Miller (WC) Elections By Age Finally, the results for the Katherine Miller (WC) Imani Williams (BL) Elections elections among income brackets are as follows. For I will now evaluate candidate choice across dif- the low-income bracket, Katherine Miller’s opponents ferent age brackets. I chose to test if those born before beat her in each race. The most significant margin was and during the Civil Rights Movement of the 1960s between Molly Harris (WL) and Katherine Miller. ascribe to stronger feelings of linked fate than those This margin was 32 percent for Katherine Miller and born after. The youngest age bracket, removed from 68 percent for Molly Harris, which is not very surpris- the Civil Rights Movement, is the age range from ing among low-income candidates. Katherine Miller twenty to thirty-three. The middle age bracket consists still lost to each of her opponents. And for the high- of those ages thirty-four to fifty-one, and the oldest income bracket, Katherine Miller still lost handily, but group consists of respondents over age of fifty-two. no p-value showed significance in any race. The tables These respondents chose Imani Williams in each elec- for the Katherine Miller elections are shown below. tion she took part of, each electing Williams at a rate of sixty percent. In each of these elections, the margin Support for Katherine Miller (White conservative) between Imani Williams and her opponent was statis- 90 the salience of race versus socioeconomic status among african american voters tically significant. Kiara Jackson (BC) Elections For the middle age group, Imani Williams still For the Kiara Jackson (BC) races, the oldest age won each race but by lesser margins, and the margins group only preferred Kiara Jackson over Katherine were not great enough to be more than minimally Miller, the white conservative, but only by a margin of significant in any of the races. For the youngest age four percentage points. Kiara Jackson lost the rest of group, Imani Williams won each race handily. The the races. Older respondents strongly preferred Molly races between Katherine Miller (WC) and Imani Wil- Harris (WL) to Kiara Jackson, voting for Molly Harris liams and Molly Harris (WL) and Imani Williams at a rate of 75 percent to 25 percent. Similarly, they were both significant. The race between Kiara Jack- preferred Imani Williams (BL), the economically lib- son (WC) and Imani Williams was not statistically eral candidate, to Kiara Jackson. significant. Imani Williams won almost 73 percent of For the middle age group in the Kiara Jackson the vote over Molly Harris (WL), and 67 percent over (BC) races, Kiara Jackson won the race only against Katherine Miller (WC). Katherine Miller (WC), by 63 percent. Kiara Jackson lost to Imani Williams, but only by a slim margin of Support for Imani Williams (Black Liberal) 53 percent to 47 percent. Similarly, Kiara Jackson lost All ages 61 percent to 39 percent to Molly Harris, the white Opponent name Vote for Imani Vote for p-value liberal. The older respondents preferred Molly Harris Williams opponent to Kiara Jackson at a rate of 75 percent to 25 percent. Katherine Miller 59% 41% 0.01771** (WC) The youngest age group in the Kiara Jackson elec- Kiara Jackson 59% 41% 0.02146** tions chose Kiara Jackson over Katherine Miller (WC) (BC) at a rate of 66 percent, and this margin was signifi- Molly Harris 62% 38% 0.001855*** cant. When Kiara Jackson was against Imani Williams (WL) (BL), respondents chose Imani Williams at a rate of 62 Born 1900-1962 percent to 38 percent. In the election of Kiara Jackson Opponent name Vote for Imani Vote for p-value Williams opponent against white liberal Molly Harris, respondents chose Katherine Miller 55% 45% 0.5465 Molly Harris 61 percent to 39 percent. Below are the (WC) results tables for the Kiara Jackson elections. Kiara Jackson 63% 37% 0.05084* (BC) Support for Kiara Jackson (Black conservative) Molly Harris 60% 40% 0.1213 (WL) All ages Opponent name Vote for Kiara Vote for p-value Born 1963-1980 Jackson opponent Opponent name Vote for Imani Vote for p-value Imani Williams 41% 59% 0.02146** Williams opponent (BL) Katherine Miller 53% 47% 0.5994 Katherine Miller 59% 41% 0.017** (WC) (WC) Kiara Jackson 53% 47% 0.6803 Molly Harris 33% 67% 0.00001963*** (BC) (WL) Molly Harris 56% 44% 0.4561 (WL) Born 1900-1962 Opponent name Vote for Kiara Vote for p-value Born 1981-1994 Jackson opponent Opponent name Vote for Imani Vote for p-value Imani Williams 37% 63% 0.05084* Williams opponent (BL) Katherine Miller 67% 33% 0.01041** Katherine Miller 52% 48% 0.7098 (WC) (WC) Kiara Jackson 62% 38% 0.1495 Molly Harris 25% 75% 0.000464*** (BC) (WL) Molly Harris 72% 28% 0.00319*** (WL) Note: * p<0.1, ** p<0.05, *** p<0.01 Values approximated to nearest whole number.

91 columbia university journal of politics & society

Born 1963-1980 Opponent name Vote for Kiara Vote for p-value Support for Molly Harris (White liberal) Jackson opponent All ages Imani Williams 47% 53% 0.6803 Opponent name Vote for Molly Vote for p-value (BL) Harris opponent Katherine Miller 63% 37% 0.07684* Imani Williams 38% 62% 0.00185*** (WC) (BL) Molly Harris 39% 61% 0.1235 Katherine Miller 62% 38% 0.002663*** (WL) (WC) Born 1981-1994 Kiara Jackson 67% 33% 0.00001963*** Opponent name Vote for Kiara Vote for p-value (BC) Jackson opponent Born 1900-1962 Imani Williams 38% 62% 0.1495 Opponent name Vote for Molly Vote for p-value (BL) Harris opponent Katherine Miller 66% 34% 0.03481** Imani Williams 40% 60% 0.1213 (WC) (BL) Molly Harris 39% 61% 0.1235 Katherine Miller 72% 28% 0.00319*** (WL) (WC) Note: * p<0.1, ** p<0.05, *** p<0.01 Kiara Jackson 75% 25% 0.000464*** Values approximated to nearest whole number. (BC) Born 1963-1980 Molly Harris (WL) Elections Opponent name Vote for Molly Vote for p-value Harris opponent In the oldest age bracket, those born between Imani Williams 44% 56% 0.4561 1900 and 1962, white Liberal Molly Harris only lost to (BL) black Liberal Imani Williams, with 60 percent of the Katherine Miller 67% 33% 0.01041** votes. Molly Harris won over both white conservative (WC) Katherine Miller and black conservative Kiara Jack- Kiara Jackson 61% 39% 0.1235 (BC) son by 72 percent and 75 percent, respectively. Both of these margins showed the significance of opponent Born 1981-1994 Opponent name Vote for Vote for p-value economic identity, meaning that older middle-class Molly Harris opponent blacks may value a specific liberal economic ideology Imani Williams 28% 72% 0.00319*** over the candidate’s race. (BL) For the middle age group in the elections involv- Katherine Miller 50% 50% 1 (WC) ing Molly Harris, Molly Harris still won against white Kiara Jackson 61% 39% 0.1235 conservative Katherine Miller and black conservative (BC) Kiara Jackson, but not quite as handily. Molly Harris Note: * p<0.1, ** p<0.05, *** p<0.01 took 61 percent of the vote against Kiara Jackson, and Values approximated to nearest whole number. 67 percent of the vote against Katherine Miller. The margin between Katherine Miller and Molly Harris Katherine Miller (WC) Elections was statistically significant. The race between Molly Finally, white conservative Katherine Miller’s Harris and black Liberal Imani Williams was fairly election results are as follows, as analyzed by age range. close, around 45 percent to 55 percent. Among the oldest age range of respondents, Katherine In the youngest age group, Molly Harris lost to Miller lost each election. The greatest margin was be- black liberal Imani Williams. Imani Williams won 72 tween her and white liberal Molly Harris, where Kath- percent to 28 percent, making this race results statisti- erine Miller took 28 percent of the vote and Molly cally significant. Molly Harris and white conservative Harris took 72 percent of the vote. This margin was Katherine Miller tied exactly, and Molly Harris won significant. Black liberal Imani Williams and black against black Conservative Kiara Jackson, gaining 60 conservative Kiara Jackson won over Katherine Miller percent of the vote. In comparison, for the two older by very slight margins. age groups, Molly Harris won in a landslide. The tables Among the middle age group of respondents, for the Molly Harris elections crossed with age bracket Katherine Miller still did not win any of the races. Be- are below. tween white conservative Katherine Miller and white 92 the salience of race versus socioeconomic status among african american voters liberal Molly Harris was the largest margin of defeat, Born 1981-1994 where Molly Harris took 67 percent of votes and Kath- Opponent name Vote for Vote for p-value Katherine Miller opponent erine took 33 percent. This margin of difference was Imani Williams 33% 67% 0.01041** statistically significant. Similar to the oldest age group, (BL) black liberal Imani Williams took only about 53 per- Kiara Jackson 34% 66% 0.03481** cent of the vote over Katherine Miller, who took 47 (BC) percent. Black conservative Kiara Jackson took about Molly Harris 50% 50% 1 (WL) 63 percent of the vote against Katherine Miller’s 37 percent. Note: * p<0.1, ** p<0.05, *** p<0.01 Finally, among the youngest age group in the Values approximated to nearest whole number. contests with Katherine Miller, Katherine Miller lost LINKED FATE AND NEIGHBORHOOD QUALITY the elections between black liberal Imani Williams RESULTS and black conservative Kiara Jackson, but there was a draw between Katherine Miller and white liberal Neighborhood Quality Questions Molly Harris. Imani Williams won by a fairly large By Income margin, enough to render the difference significant. And Kiara Jackson also won handily against Kather- The eight questions that I chose to analyze ine Miller, with 66 percent of the vote, and the results regarding neighborhood demographics and quality in were also significant. Both Katherine Miller and Molly the survey were taken from the National Black Election 62 Harris received 50 percent of the vote from the young- Panel Survey from 1988. The choice of questions est group of respondents. The results for the Katherine was inspired by Claudine Gay’s 2004 article “Putting Miller elections analyzed by age are shown below. Race in Context: Identifying the Environmental Determinants of Black Racial Attitudes,” which Support for Katherine Miller (White Conservative) concerns the relation between neighborhood quality 63 All ages and make-up to linked fate. The questions were Opponent name Vote for Vote for p-value each analyzed across income groups by high, middle Katherine Miller opponent and low-income respondents. The first question Imani Williams 41% 59% 0.01771** asked was a question about the racial makeup of (BL) the respondent’s neighborhood. The answers to this Kiara Jackson 41% 59% 0.017** (BC) question were “all black,” “mostly black,” “half black,” Molly Harris 38% 62% 0.002663** and “less than half black.” For the high-income group, (WL) about 53 percent of respondents said that they lived Born 1900-1962 in neighborhoods that were half or more black, while Opponent name Vote for Vote for p-value about 47 percent said they live in neighborhoods that Katherine Miller opponent are less than half black. In the middle-income group, Imani Williams 45% 55% 0.5465 (BL) 64 percent of respondents live in half or more black Kiara Jackson 48% 52% 0.7098 neighborhoods, while 36.25 percent of middle-class (BC) respondents live in neighborhoods with less than half Molly Harris 28% 72% 0.00319*** blacks. The percentages for “mostly black” and “less (WL) than half black” 30 and 36 percent respectively, are Born 1963-1980 similar in the case of middle-income respondents. Opponent name Vote for Vote for p-value Katherine Miller opponent The next question will show what perceived “class” of Imani Williams 47% 53% 0.5994 neighborhood each respondent lives in. The lower- (BL) income respondents answered that 67 percent live in Kiara Jackson 37% 63% 0.07684* half or more black neighborhoods, and 33 percent live (BC) in less than half black neighborhoods. The low-income Molly Harris 33% 67% 0.01041** (WL) group has the highest percentage of respondents that live in both all black and mostly black neighborhoods, but not by a very large margin. For the low-income blacks, the percentage of those who live with less than 93 columbia university journal of politics & society half blacks is surprisingly not much lower than that Neighborhood class status makeup percentage for the middle class. Income Poor Working Middle Upper Upper I don't class class middle class know Neighborhood racial make-up class Income All black Mostly black Half black Less than High 2% 29% 46% 19% 2% 2% half black Middle 3% 45% 39% 9% 1% 2% High 7% 26% 20% 47% Low 10% 49% 30% 5% 1% 6% Middle 10% 30% 24% 36% Low 11% 32% 23% 33% Values approximated to nearest whole number. Values approximated to nearest whole number. The second question in the neighborhood The next set of neighborhood questions section asks about the socioeconomic status of a asked about the quality of different resources and respondent’s neighborhood. The answers to this institutions in the respondent’s neighborhood. This question include “poor,” “working class,” “middle- set included questions about police protection, road class,” “upper middle-class,” “upper-class,” and “I don’t maintenance, parks and playgrounds, public schools, know.” The results across all incomes are as follows. and garbage collection. The first question asked Respondents from poor neighborhoods made up about the quality of the police protection in each 4.19 percent of the total respondents. Those from respondent’s neighborhood. The answer choices were the working class totaled 33.68 percent. In turn, “very dissatisfied,” “somewhat dissatisfied,” “somewhat 30.68 percent responded that they were middle class, satisfied,” and “very satisfied.” The high-income and 8.46 percent said that they were upper middle respondents were, on the whole, happier with their class. Upper-class neighborhood respondents made neighborhood’s police protection than low-income up 0.94 percent of the total respondents, and those respondents. who didn’t know their neighborhood’s social class responded 19.32 percent of the time. The analysis for Satisfaction with neighborhood police protection Income Very Somewhat Somewhat Very those of high-income showed that about 31 percent dissatisfied dissatisfied satisfied satisfied lived in lower class neighborhoods, 46.26 percent High 5% 12% 54% 28% lived in a middle-class neighborhood, and about 21 Middle 10% 15% 52% 23% percent lived in upper class neighborhoods. In sum, Low 9% 19% 52% 20% high-income respondents said that they lived in Values approximated to nearest whole number. middle-class neighborhoods more than any type of neighborhood. For the middle class, almost 50 percent The next question regarding neighborhood said they live in lower-class neighborhoods, and 39.50 quality dealt with residents’ satisfaction with percent responded that they lived in a middle-class neighborhood parks and playgrounds. The answer neighborhood. About 10 percent lived in upper-class choices were the same as above. In terms of the neighborhoods. 1.88 percent of respondents said they highest income voters, about 18 percent were did not know what class of neighborhood they lived unhappy with the park and playground quality in in. In conclusion, most middle-income respondents their neighborhoods, while about 82 percent were said they live in lower class neighborhoods. satisfied with their neighborhood parks. Overall, the For low-income respondents, almost 60 percent high-income respondents were very pleased with the said they live in lower-class neighborhoods, whereas 30 status and quality of their local parks and playgrounds. percent said they live in a middle-class neighborhood, For the middle-income respondents, about 32 percent and about 6 percent said they live in more upper-class were unhappy with the state of their parks. About 68 neighborhoods. 5.8 percent of respondents did not percent were happy with their neighborhood’s park know what kind of neighborhood they live in. and playground quality. And as for the low-income respondents, about 31 percent were dissatisfied, and about 69 percent were happy with park quality. The middle- and lower- class respondent answers were similar, and if anything, the middle class was more satisfied. The upper-class respondents were almost 15 94 the salience of race versus socioeconomic status among african american voters percentage points happier with their neighborhood Satisfaction with neighborhood health clincs parks and playgrounds. Income Very Somewhat Somewhat Very dissatisfied dissatisfied satisfied satisfied High 6% 10% 51% 34% Satisfaction with neighborhood parks and playgrounds Middle 5% 14% 41% 40% Income Very Somewhat Somewhat Very dissatisfied dissatisfied satisfied satisfied Low 8% 16% 45% 32% High 6% 13% 49% 32% Values approximated to nearest whole number. Middle 10% 22% 40% 28% Low 13% 18% 46% 23% The last neighborhood quality question in Values approximated to nearest whole number. the survey was about respondent satisfaction with neighborhood garbage collection. Again, the answer The next question regarding neighborhoods choices are the same as above, and the overall asked about respondent satisfaction with percentages for all income levels are as follows. Overall neighborhood public schools. The answer choices dissatisfaction with garbage collection is 16 percent, were the same as the about neighborhood quality whereas general satisfaction is 83 percent. In the questions. For the high-income bracket, around 28 high-income bracket, the dissatisfaction is around 10 percent of respondents were dissatisfied with their percent, and the satisfaction is around 90 percent. For public schools, and 72 percent were satisfied. For the the middle-income respondents, the dissatisfaction middle-income bracket, the figures were fairly similar, is around 17 percent, and the satisfaction is near 83 29 percent and 71 percent, respectively. And among percent. And among low-income respondents, the the low-income respondents, the satisfaction was overall dissatisfaction is around 23 percent, and the about the same: 30 percent were dissatisfied, whereas satisfaction is around percent. Not surprisingly, the 70 percent disagreed. low-income respondents have the lowest-quality garbage collection services, and the high-income Satisfaction with neighborhood public schools respondents have the highest quality. Income Very Somewhat Somewhat Very dissatisfied dissatisfied satisfied satisfied Satisfaction with neighborhood garbage collection High 9% 19% 44% 28% service Middle 13% 16% 44% 27% Income Very Somewhat Somewhat Very Low 11% 19% 45% 25% dissatisfied dissatisfied satisfied satisfied Values approximated to nearest whole number. High 5% 5% 38% 52% For the next question, which gauged satisfaction Middle 3% 10% 37% 50% with respondent neighborhood health clinics and Low 6% 10% 40% 44% hospitals, the answers were the same as the above. Values approximated to nearest whole number. For high-income respondents, 15 percent were dissatisfied with health clinics and hospitals in their Overall, we see that the low-income respondents neighborhoods, and 85 percent were satisfied. For have the lowest-quality resources available in their the middle-income respondents, these numbers neighborhoods, while higher-income respondents were 18% dissatisfied and 81 percent satisfied. And have access to higher-quality resources and are on lastly, for low-income respondents, 23 percent were average more satisfied with their neighborhood quality. unhappy with their local healthcare, and 76 percent This satisfaction, or lack thereof, has implications for were satisfied. There is a small jump from low-income strength of linked fate among each socioeconomic to high-income dissatisfaction, from 15 percent to class. 23 percent. This reveals that the quality of healthcare in low-income black neighborhoods may be slightly Linked Fate Questions lower than in high-income neighborhoods. By Income The questions on linked fate come from the 1988 Black National Election Panel Survey.64 These questions ask about the extent to which black respondents ascribe to Dawson’s notion of linked fate 95 columbia university journal of politics & society and African American group consciousness. I chose Agree or disagree: In the United States, if black people eight questions to analyze by both income and age. The don't do well in life it is because they are kept back three income groups are those used previously. The because of their race. age brackets are the same brackets described above Income Agree Agree Disagree Disagree Neither strongly somewhat somewhat strongly agree nor as well. First, I will analyze the linked fate-themed disagree questions by income bracket. The first question asks High 14% 42% 26% 13% 4% black respondents if they think that what happens to Middle 13% 36% 25% 17% 10% blacks in the United States affects them. The answers Low 15% 38% 21% 16% 9% are “yes, a lot,” “yes, some,” “yes, not very much,” and Values approximated to nearest whole number. “no, not at all.” I first looked at the answers for all the respondents. For the high-income bracket, those that The third question asks if blacks as a whole are agree with this description of linked fate and whose getting along well, and to what degree. The answer answers were “yes, a lot” or “yes, some”, make up 73 choices for this question are “very well,” “fairly well,” percent of the total high-income bracket. For the “not too well,” and “not well at all.” Again, I group the middle-income group, those who agree that what two “well” choices together and the “not well” choices happens to other blacks affects them, make up 70% together to get a more comprehensible depiction of of the bracket, and those who don’t really agree make respondent answers. For those in the high-income up 30%. And among low-income voters, 73% have a group, 60 percent believe that blacks are getting sense of linked fate, and 27% do not. along well, and 40 percent believe the opposite. For middle-income blacks, 62 percent believe that blacks Do you think that what happens generally to black are getting along well, and 38 percent believe that they people in this country will have something to do with are not. And among low-income voters, 63 percent what happens to your life? believe that blacks are getting along well, and 37 Income Yes, a lot Yes, some Yes, not very No, not at all much percent believe that they are not. High 28% 45% 13% 14% Middle 31% 39% 12% 18% Would you say that blacks as a group are getting along Low 32% 41% 11% 16% very well, fairly well, not too well or not well at all? Income Very well Fairly well Not too well Not well at Values approximated to nearest whole number. all High 10% 50% 35% 5% The second linked fate question asks respondents Middle 9% 53% 31% 6% to answer whether they think that if black people don’t Low 13% 50% 28% 9% do well in life, it is because of their race. The answer Values approximated to nearest whole number. choices for this question were “agree strongly,” “agree somewhat,” “disagree somewhat,” “disagree strongly,” The fourth question asks if respondents think and “neither agree nor disagree.” For the high-income that blacks as a group are getting along economically. section, 56 percent agree that blacks are kept back The answer choices are the same as above. Among by their race, 39 percent disagree, and 4.3 percent the high-income respondents, 43 percent said that neither agree nor disagree. The middle-income blacks are getting along well, and 57 percent said that section answered that 48 percent agree, 42 percent blacks are not getting along well. The middle-income disagree, and 9.6 percent did not agree or disagree. group responded 43 percent of the time that blacks Interestingly, the middle class seems to ascribe less to are getting along well, and 57 percent said that they this notion of linked fate than does the upper class, by are not. The low-income respondents answered that 8 percentage points. Among the low-income voters, 49 percent think blacks are doing well, and 51 percent 53 percent agree, 38 percent disagree, and 9 percent answered that they are not. are neutral. So, the high-income respondents show the highest degree of linked fate for this question, then low-income respondents, and the middle-income voters show the lowest degree.

96 the salience of race versus socioeconomic status among african american voters Would you say that blacks as a group are getting along exactly the same. And for low-income respondents, economically very well, fairly well, not too well or not 72 percent of respondents believe that the black well at all? Movement affected them, while 27 percent believe Income Very well Fairly well Not too well Not well at that the Movement did not affect them. all High 6% 37% 48% 10% Do you think that the Movement for black rights has Middle 8% 35% 45% 12% affected you personally, and if so, how much? Low 10% 38% 39% 13% Income A lot Some Not very Not at all Values approximated to nearest whole number. much High 32% 42% 20% 6% The fifth question asks again about the economic Middle 29% 44% 17% 9% position of blacks, but in comparison with that Low 33% 40% 17% 10% of whites. The five choices for this question were: Values approximated to nearest whole number. “much better,” “somewhat better,” “about the same,” “somewhat worse,” and “much worse.” Overall, 19 The next question asks if respondents feel close to percent of respondents think the economic position other blacks, and if so, how close they feel. The answers of blacks is better than whites, 21 percent said the are “very close,” “fairly close,” “not too close,” and “not position is the same, and 59 percent said that blacks are close at all.” Again, I group responses by generally worse off economically than whites. For high-income feeling close versus not generally feeling close. Eighty- respondents, 18 percent said that the black position five percent of upper class respondents feel close to is better, 17 percent said that the black position is the other blacks, while 14% do not feel so close to other same, and 65 percent said that the black position is blacks. Eighty percent of middle-class respondents worse than whites economically. Among the middle- feel that they are close to other blacks, while 20 percent income respondents, 20 percent said that the black do not feel close to other people of their race. Eighty- position is better, 20 percent said it is the same, and three percent of lower-class respondents feel close to 60 percent said that blacks are worse off than whites in other blacks, while 18 percent do not feel as close. the fiscal realm. And for low-income respondents, 20 percent said the black position is better than whites, Do you feel very close, fairly close, not too close, or 26 said the position of blacks is the same as whites, not close at all to black people in this country? and 54 percent say that the black financial position is Income Very close Fairly close Not too Not close at close all worse than whites. High 33% 53% 13% 1% Middle 33% 47% 18% 2% On the whole, would you say that the economic position of blacks is better, about the same, or worse Low 37% 45% 15% 3% than whites? Values approximated to nearest whole number. Income Much S ome w hat About Somewhat Much better better the same worse worse The final linked fate question asked respondents High 8% 11% 17% 42% 23% about a hypothetical mayoral election between two Middle 5% 14% 20% 38% 22% black candidates. The question asks what characteristic Low 6% 14% 26% 35% 19% of one candidate would make the respondent vote Values approximated to nearest whole number. for that candidate over their opponent. The options were: “qualification/experience,” “stand on issues,” The next question is whether the Movement for “honesty/integrity,” “cares for all people/whole city,” black rights has affected respondents. The answers to “cares for black people first,” “competence/leadership,” the question are “a lot,” “some,” “not very much,” and “personality/charisma,” “Democrat,” “supports equal “not at all.” The results for the high-income class are opportunity/no discrimination,” “cares for poor/ as follows: 74 percent of respondents believe that the homeless,” and “other (please specify).” High-income Movement did affect them, while 26 percent did not respondents answered that they would choose the really feel that the Movement affected them. For the candidate based on their stand on issues 18 percent middle-income respondents, the percentages were of the time, and 17 percent of the time they would 97 columbia university journal of politics & society vote for the candidate who shows honesty/integrity. Do you think that what happens generally to black Sixteen percent of the time high-income respondents people in this country will have something to do with would vote for the candidate who cares for the whole what happens to your life? city, and only 0.7 percent of the time would high- Age Yes, a lot Yes, some Yes, not very No, not at all much income respondents vote for the candidate who puts Elder 34% 41% 11% 15% black concerns first. Eleven percent of high-income Middle 30% 41% 12% 17% respondents would vote for the Democratic candidate. Middle-income respondents had very much the same Young 28% 43% 14% 16% ages for the above listed vote-choices. For low-income Values approximated to nearest whole number. respondents, the percentages also did not vary much. However, 2% of respondents of low-income would For the question: “In the United States, if black vote for the candidate who cares about black people people don't do well in life it is because they are kept first, but 18% would vote for those who care most for back because of their race”, respondents were asked the people’s and the city’s welfare. how strongly they agree with this statement. Analyzing for age groups, the oldest age group responded that Income Qualification/ Stand Honesty/ Cares Cares 54% agree with this definition of linked fate, 39 experience on integrity for all for black percent of them did not agree, and 7 percent neither issues people/ people whole first agreed nor disagreed. In the middle age group, 52 city percent agreed, 38 percent disagreed, and 10 percent High 9% 18% 17% 16% 1% neither agreed nor disagreed. And for the youngest Middle 13% 18% 17% 17% 2% age group, 51 percent agreed that if blacks don’t do Low 6% 17% 15% 18% 2% well, it is because of race, and 42 percent did not agree Competence/ Person- Democrat Equal Cares Other with this statement. Seven percent neither agreed nor Leadership ality/ opportunity/No for disagreed. Although these numbers are fairly similar, Char- discrimination poor/ isma Home- the oldest group, born before and during the Civil less Rights Movement, has the highest degree of linked 14% 2% 11% 8% 3% 0% fate, though only by 3 percentage points. 9% 2% 10% 9% 4% 0% 9% 1% 15% 10% 6% 1% Please indicate whether you agree or disagree with Values approximated to nearest whole number. the following statement: In the United States, if black people don't do well in life it is because they are kept By Age back because of their race. For the first question on linked fate, asking about Age Agree Agree Disagree Disagree Neither strongly somewhat somewhat strongly agree not how much respondents feel that what happens to other disagree blacks will affect them, 48 percent of older respondents Elder 14% 40% 27% 12% 7% ascribed to this sense of linked fate, while 52 percent Middle 13% 39% 21% 17% 10% did not really identify with it. For the middle-income Young 13% 38% 23% 18% 7% group, 47 percent of respondents said that they Values approximated to nearest whole number. believed in this kind of linked fate, while 53 percent did not really feel it too much. And for the youngest The question that asks if blacks as a group are group, 43 percent of respondents said that they feel getting along well yielded the following results in linked to other blacks, while 57 percent said that they the test across age brackets. For the oldest bracket, do not feel this type of linked fate. It is interesting respondents said 45% of the time that blacks are doing to note that around 40 percent of respondents in well, and 55 percent of the time that blacks are not all age groups answered that they felt “some” linked doing well. For the middle age group, 41 percent of fate. However, about a third of the oldest bracket of respondents said that blacks as a group are getting respondents said that they felt “a lot” of linked fate, along well, and 59 percent said that blacks are not compared to 30 percent and 28 percent of middle age getting along well. And for the youngest group, 49 and lower age bracket respondents, respectively. percent of respondents said they feel that blacks 98 the salience of race versus socioeconomic status among african american voters are getting along well in the United States, while 51 about the same, and 60 percent feel that blacks are percent said that blacks are not getting along as well. worse off. Again, these numbers are not too different, although the older group agrees by 4 percent more Would you say that blacks as a group are getting along than younger age groups that blacks are financially very well, fairly well, not too well or not well at all? worse off than whites. Age Very well Fairly well Not too well Not well at all On the whole, would you say that the economic Elder 4% 41% 44% 11% position of blacks is better, about the same, or worse Middle 10% 31% 46% 13% than whites? Young 11% 38% 40% 11% Age Much Somewhat About the Somewhat Much Values approximated to nearest whole number. better better same worse worse Elder 3% 13% 20% 45% 19% The next question is similar to the one above, Middle 6% 14% 20% 35% 25% but asks if respondents think blacks are economically Young 9% 13% 24% 35% 19% doing well. For those born between 1900 and 1962, Values approximated to nearest whole number. 45 percent said that blacks are getting along well economically, and 55 percent said that blacks are not The next question is very salient for whether and getting along well in the financial realm. Those in the how much each age group feels that the black rights middle age realm say 41 percent of the time that blacks Movement affected them personally. The existing are getting along well economically, and 59 percent would suggest that the older group to respond that of the time that black and not getting along so well they were more affected by the Movement than the economically. And in the young age group, those born younger respondents. The oldest group responded between 1981 and 1994, 49 percent of respondents say that 72 percent felt affected by the Movement, and that blacks are fine fiscally, while 51 percent do not 28 percent did not feel as affected by it. Among the feel this way about the economic situation of the black middle age respondents, 70 percent said they felt group. affected by the Movement, and 30 percent did not feel very affected by the Movement for black rights. And Would you say that blacks as a group are getting along for the lowest age group, 79 percent said that they felt economically very well, fairly well, not too well or not affected by the Movement, and 21 percent did not feel well at all? affected by the Movement. Age Very well Fairly well Not too well Not well at all Do you think that the Movement for black rights has Elder 4% 41% 44% 11% affected you personally, and if so, how much? Middle 10% 31% 46% 13% Age A lot Some Not very Not at all Young 11% 38% 40% 11% much Values approximated to nearest whole number. Elder 29% 43% 21% 7% Middle 28% 42% 18% 12% The next question concerns a similar topic, but Young 38% 41% 15% 6% asks if this economic position is worse, better, or the Values approximated to nearest whole number. same as whites. In the oldest age range, 16 percent of respondents feel that the economic position of The next question asks respondents how close blacks is better than whites, 20 percent feel that the they feel to other blacks in the country. For the oldest position is about the same, and 64 percent believe that age group, 85 percent say that they feel close to other economically, blacks are worse off than whites. For the blacks, and 15 percent say they do not feel as close to middle age group, 20 percent of respondents believe other blacks. For the middle age group, 82 percent that blacks are better off than whites, 20 percent said they feel close to other blacks, and 18 percent said believe they are about the same, and 60 percent believe they do not feel as close to other blacks. Conversely, that blacks are worse off than whites economically. In 79 percent of young blacks say that they feel close to the youngest age bracket, 22 percent feel that blacks other people of their race, and 21 percent do not feel are better off economically, 24 percent feel they are as close. The oldest age group seems to feel more of a 99 columbia university journal of politics & society connection to other blacks than do other age groups Competence/ Personality/ Democrat Equal Cares Other Leadership Charisma oppor- for poor/ of African Americans. tunity/No Home- discrim- less Do you feel very close, fairly close, not too close, or ination not close at all to black people in this country? 12% 0% 13% 9% 5% 0% Age Very close Fairly close Not too Not close at 11% 1% 13% 9% 3% 1% close all 9% 4% 11% 9% 4% 0% Elder 41% 45% 13% 1% Values approximated to nearest whole number. Middle 31% 51% 15% 3% Young 31% 48% 17% 4% DISCUSSION Values approximated to nearest whole number. By Income What can be concluded about different income The final linked fate question about the and age groups of African Americans? For the low- hypothetical election between the two black candidates income group, overall, it seemed that economic policy was analyzed in age brackets. The results are as follows. was more important than just the perceived race of For the eldest age group, those respondents born the candidate. Among these low-income respondents, between 1900 and 1962, just 0.3 percent answered Molly Harris, the white liberal, won handily over Kiara that they would vote for the candidate who puts black Jackson, the black conservative. Under the idea of racial people first, while 18 percent said that they would linked fate, these results would not hold. Nevertheless, vote for the candidate who cares about the whole economic linked fate seems to be tying together city. Thirteen percent said they would vote for the the lower class of African Americans. However, in Democratic candidate. For the age group born between accordance with Dawson’s explanation of the “new 1962 and 1980, 1 percent said that they would vote Black Politics” as emerging from the roots of slavery for the candidate who prioritizes black issues, while and economic oppression, the black community has 16 percent said they would choose the candidate who historically experienced a lower economic status.65 takes the whole city’s needs into consideration. Twelve Thus, that African Americans of low-income voted for percent of the middle age group said they would vote a white candidate with a liberal fiscal policy is perhaps for the Democrat. In the youngest age bracket, 3.6 not so surprising. It would most likely go against percent said they would vote for the candidate who the majority of the black group to vote for a fiscally takes on black needs first, and 17 percent said that conservative candidate as Kiara Jackson, even though they would choose the one who looks at the needs of she is black. Voting for the liberal candidate, no matter the whole city. Eleven percent said they would vote for what the race, seems to still be aligned with the black the Democratic candidate. political community. Low-income respondents are generally more If there were only two black candidates running likely to view the black position as slightly better against each other for mayor in your city or town, than both the middle and high-income groups view what would be the most important factor that would it. The level of linked fate among low-income blacks make you vote for one over the other? could result from both being around only African Age Qualification/ Stand on Honesty/ Cares Cares Americans and thus not feeling like an out-group, experience issues integrity for all for black people/ people in addition to low-income neighborhoods being less whole first socially and politically active and thus potentially less city united in these areas. Elder 12% 13% 17% 18% 0% About two-thirds of low-income respondents Middle 7% 21% 17% 16% 1% live in half or more black neighborhoods, about Young 9% 19% 14% 17% 4% 60 percent say they live in lower or working class neighborhoods, and 30 percent say they live in middle-class neighborhoods. In terms of satisfaction with neighborhood amenities, low-income blacks are happy enough with their police protection, parks and playgrounds, neighborhood schools, healthcare, and 100 the salience of race versus socioeconomic status among african american voters garbage collection. Though the average satisfaction fate may depend on where they work, or where on the rate was a bit lower than that of middle-income income spectrum they lie. respondents, the lowest percentage of satisfaction Middle-income respondents said that 64 was 69 percent. Gay writes that, “In lower-status percent of them live in neighborhoods that are half neighborhoods, where the absence of economically or more black. Almost half said that they lived in stable and secure families may erode a community’s lower class or working class neighborhoods, while organizational strength, the dearth of informal about 40 percent said they lived in a middle-class and formal institutions leaves residents largely neighborhood. Middle-income respondents are disconnected from the ‘networks of communication generally satisfied with their neighborhood police and community learning’ (Cohen and Dawson 1993, protection, public schools, hospitals and health clinics, 290) so critical to the development and diffusion of and garbage collection services, with the lowest rate of African Americans.”66 satisfaction being 68 percent. Again, the middle class Lower income respondents thus may have a finds itself in the middle of black ideologies. They lower degree of black group unity, but above discussed are generally satisfied with the available resources, is the high economic unity stressed by low-income but they live in mostly lower-class neighborhoods African Americans. Indeed, though, the low-income around other blacks. Thus, they may keep company respondents do not seem to display as strong of a with both middle- and low-income blacks, which racial group consciousness. This may also be because most likely increases a sense of shared fate. However, they live around mostly other African Americans, and Gay writes that, “African Americans who enjoy access so don’t feel as singled out as a higher status blacks to the resources and opportunities available in better might. neighborhoods may identify only weakly with the Middle-income voters are in a unique history of racial suffering that underlies notions of position within the African American community. shared fate and predisposes blacks to view race as the Because they are on an economic rise, voting for defining interest in their lives.”67 As a result of these fiscally liberal policies may not always benefit or suit two contrasting environmental factors, it is difficult to them. However, being African-American may make predict how middle-class blacks will act politically. middle-class black voters more adherent to liberal For middle-income respondents, it seems that fiscal policies. The middle class voted more strongly linked fate is not quite as strong. For the first few linked in favor of Imani Williams over Kiara Jackson than fate questions, middle-income respondents ascribed did both the lower and upper classes. In addition, in to the lowest degree of linked fate of both low and the race between Kiara Jackson and Molly Harris, high-income respondents. In addition, they feel the middle-class respondents voted for Molly Harris least close to other blacks of all three income groups. nearly as strongly as did the lower class. However, the Still, middle-income respondents feel as connected to split between the votes for Katherine Miller versus the black group as do high-income respondents, and Kiara Jackson among the middle class was low— agree in a couple more areas with the high-income Kiara Jackson won by just 55 percent. The middle respondents. It is difficult thus to pin down middle- class still preferred the black conservative over the income blacks, because of the countering forces of white conservative, but not by much. And in the race environmental factors. The sample of middle-class between the white conservative and white liberal, the blacks in my study seemed to be more on the lower- results were 60 percent to 40 percent for the white middle-class end of the spectrum but were fairly liberal, but these were around the same%ages as the satisfied with the quality of their neighborhoods. high-income respondents. In the first two races, the In addition, they lived mostly among other blacks. middle class respondents seemed to have the highest However, there was a fairly broad spectrum of incomes sense of linked fate of all the income groups, but in included in this subset, and there was a substantial the second two cases, it seems like the middle class portion of middle-class blacks that lived among a shows less concern for the black group as a whole. The minority of blacks. Perhaps these two ends of the middle class takes on both characteristics of the upper middle-class spectrum had differing levels of linked class and the lower class. Perhaps depending on where fate, and thus cancelled each other out on some of a middle-class respondent lives, their sense of linked the survey questions. The fact that most middle-class

101 columbia university journal of politics & society respondents live amongst other blacks may increase as strong a sense of linked fate as do low-income sense of linked fate, but the fact that middle-class respondents. These respondents also have the highest neighborhood resources are acceptable would negate tendency to believe that the black group as a whole is this sense of linked fate. This is somewhat seen in the not doing well. High-income respondents also believe ambiguity of middle-class results, though there still that blacks as a whole are worse off economically, over exists a fairly strong sense of linked fate among middle- ten percentage points more so than do low-income class respondents overall; it is just not as consistently respondents. Those of high-income also say they feel shown as either the upper or lower classes. the closest to other blacks, closer than do both middle Among the high-income respondents, trends and low-income respondents. included a seemingly high sense of linked fate. In conclusion, linked fate seems to be a High-income respondents voted for Imani Williams salient factor among black voters of all incomes, but over Kiara Jackson at about the same rate as did seems highest among high-income voters. This could low-income respondents. This seems economically confirm Claudine Gay’s hypothesis that, “African counterintuitive for high-income respondents, but Americans in neighborhoods with high-status black perhaps this group still identifies strongly, in racial residents are more likely than African Americans terms, with the black group. Otherwise, perhaps in low-status neighborhoods to view race as the these economic policies might be helpful for the part defining interest in their lives.”70 Thus, those high- of the spectrum that makes closer to $60,000, rather income African Americans who live amongst other than the end that makes $100,000. In addition, the high status blacks are likely to have a stronger sense high-income respondents chose Molly Harris (WL) of linked fate. This is because these higher status over Kiara Jackson (BC), even if only by a statistically neighborhoods often have more group activities that insignificant margin. The high-income respondents unite the community and foster the sharing of ideas, also preferred Imani Williams over Katherine Miller, and lower income neighborhoods are less likely to by a 61 percent margin, demonstrating again that race have these institutions.71 However, the reason for may be more important to high-income black voters strong linked fate among high-income blacks may than economic policy. be due to discrimination faced by inhabiting mostly On the whole, high-income respondents white areas and frequently interacting with whites. were almost split on the number of blacks in their High status blacks may be more willing to ascribe neighborhood, although just over half said that they to the black group because they feel a stronger sense were in mostly black neighborhoods. High-income of group consciousness and unity with other blacks respondents also reported that almost 50 percent of than among whites.72 These explanations involve them lived in middle-class neighborhoods. High- the environmental reasons for linked fate, which income respondents were happy with the police are discussed further below in the neighborhood protection, parks and playgrounds, public schools, demographics and quality section of the analysis. health clinics, and garbage collection services at rates of 72 percent or higher. Claudine Gay writes that By Age since this sector of the black community has access For the oldest age group, around half of to such good neighborhood resources, they should respondents said they did ascribe to linked fate, and share a weaker sense of linked fate with the rest of the over half said that blacks are not getting along well black community.68 Gay posits, “African Americans in economically. The oldest age group is also the most neighborhoods with few amenities are more likely than likely to believe that blacks are worse off economically African Americans in high-quality neighborhoods to than whites. Interestingly enough, the youngest age view race as the defining interest in their lives.”69 This group expressed that the Movement for black rights aspect of Gay’s hypothesis suggests that high-income affected them personally, more so than did the oldest African Americans will show a weaker sense of shared age group. I had hypothesized that older blacks fate with other blacks. But, there are reasons to believe would have felt more moved by the rights Movement. that high-income African Americans show a strong Nevertheless, my survey questions did not specify sense of shared destiny with African Americans. which particular Movement for black rights was High-income respondents tend to have about referred to. If the question had specified the Civil

102 the salience of race versus socioeconomic status among african american voters

Rights Movement of the 1960s, perhaps responses comes from a background of economic subjugation.73 may have been different. Still, those born between Thus, African Americans of the older age groups may 1900 and 1962 demonstrate high levels of linked fate be more in tune with the liberal economic policy. in general, and perhaps some of this could be linked Still, it is interesting that the younger group of voters to the fact that these respondents grew up before and had such strong racially linked fate. However, the during the Civil Rights Movement. Clemetson article referred to above discusses the race The middle age group had high linked fate, of a black Republican, Michael Steele, winning 30 but did not take a clear side on whether it thought the percent of the black vote in Baltimore by making radio black group is getting along well. This group also felt ads on the Baltimore hip-hop stations.74 This black the least affected by the Movement for black rights, political leader’s appeal to young blacks is somewhat even though most in the age bracket were born during similar to what is seen in the younger contingency of or right after the Movement. Overall, the middle age survey respondents in the hypothetical elections, as group had high levels of linked fate, though the oldest they voted at fairly high rates for the black candidate group was slightly higher. no matter what fiscal policy they supported. The youngest age group showed slightly lower levels of linked fate and connectedness to the black CONCLUSION group overall, but, as mentioned above, those in the It can still be argued that because there was not an eighteen to twenty-five age bracket said they felt most overwhelming amount of significant difference overall moved by the Movement for black rights. In addition, in strength of linked fate between incomes or age 3.5 percent of young respondents said they would vote groups among the black community, black politicians for the black candidate who cares for black people can still appeal to a fairly unified black political group. first, while the other age group chose this option less All of the age and income groups ascribed to high than 1 percent. Perhaps younger black voters are more levels of linked fate, and demonstrated that being likely to ascribe to a black candidate, but do not really African American is still more important than one’s feel as close to other blacks as do older blacks. economic status when choosing a political candidate. Concluding from the results of the Still, emphasizing economic liberality seems to be hypothetical elections as analyzed by age, the older an important facet of the black vote, regardless of two age groups showed more economic linked fate, race. We saw that the high-income respondents still while the youngest group in fact showed more racial strongly preferred a white liberal candidate to a black linked fate. This was interesting, as I had expected that conservative candidate, one who in fact matches along the groups born before and during the Civil Rights both racial and fiscal lines. All things considered, the Movement would have had more racial linked fate, black group seems to be a solidly unified and unique meaning that they would have voted for the candidate political unit with a tendency toward economic who was black. However, instead, these two groups liberality. voted at high rates for the candidates who exhibited Some parts to further the scope of this project the liberal fiscal policy, with the oldest group voting would be to analyze survey questions across education at even higher rates than the middle group. The level as well as neighborhood quality. Claudine Gay youngest group was the most adamant about voting analyzed the neighborhood makeup via education for the black candidate, and voted for Kiara Jackson level in her 2004 article, and Gay also wrote about and Imani Williams at higher rates than both the the effects of neighborhood quality and resources on older groups. For the oldest group, it didn’t matter one’s sense of linked fate.75 I hinted at the potential as much whether the liberal candidate was black that neighborhood quality might have on linked or white, though the black liberal still won by a 20 fate, but did not run a cross-tabulation with my own percent margin. What seemed to have mattered most data to test this hypothesis. There are several other was that the candidate was economically liberal. Like cross-tabulations that would bring more light to the the lowest income group analyzed above, it seems that question of the salience of race versus socioeconomic economic policy was more important in the elections status, such as crosses with perceived social class, and than just the candidate’s race, likely because of zip code. In addition, visiting the four New Jersey Dawson’s explanation of the black group as one which townships directly and conducting an in-person,

103 columbia university journal of politics & society qualitative study on resources and conditions of each Political and Social Research, 1997. Pew Research Centers Social Demographic Trends Project, “Blacks See town would inform the neighborhood/linked fate Growing Values Gap Between Poor and Middle Class." correlation as well. With the dataset, there is much http://www.pewsocialtrends.org/2007/11/13/blacks-see-growing-values-gap- between-poor-and-middle-class/ more that can be learned about African Americans, Kennedy, Theresa R. "Does Race Trump Socioeconomic Status in Shaping linked fate, and politics. the African American Vote?" Junior Paper (Unpublished). Princeton University, 2013. In conclusion, it remains to be seen whether Gay, Claudine. "Putting Race in Context: Identifying the Environmental the monolithic black group will ever break up. It has Determinants of Black Racial Attitudes." American Political Science in some ways, but the core of the African American Review 98.04 (2004): 547-562. Schuessler, A. A. (1999). Ecological inference. Proceedings of the National political group remains strong. Dawson writes in his Academy of Sciences, 96(19), 10578-10581. 2001 book, Black Visions: The Roots of Contemporary Dawson, Michael C. Black Visions: The Roots of Contemporary African- American Political Ideologies. Chicago: University of Chicago, 2001. African-American Political Ideologies: Print. On issues of taxes, partisanship, the role of The New York Times. "Younger Blacks Tell Democrats to Take Notice." http:// www.nytimes.com/2003/08/08/national/08VOTE.html. government, fiscal policy, and the like, blacks A&E Television Networks. "Civil Rights Movement." http://www.history.com/ remain on the left and unified—more unified topics/black-history/civil-rights-movement. across class than whites, but on issues of strategy, The New York Times "Black and Bruised." http://www.nytimes.com/2004/02/01/ tactics, and norms of the black quest for social magazine/01VOTERS.html. Gay, Claudine and L. Hochschild, Jennifer. “Is Racial Linked Fate Unique? justice, large cleavages can be detected even Comparing Race, Ethnicity, Class, Gender, and Religion”. Paper using the crude instrument of the public opinion prepared at the Annual meeting of the American Political Science survey.76 Association, 2010 ArcGIS. "Stanford Map of 2008 Precinct Winners." Arguably, some of the “institutions and networks ABC News. "Top 20 'Whitest' and 'Blackest' Names.". http://abcnews. go.com/2020/story?id=2470131. within the black community that reinforce racial Mongabay.com. "Most Common Last Names for Whites in the U.S.". http:// schema,” which Dawson discusses in his 2001 book, names.mongabay.com/data/white.html. Mongabay.com “Most Common Last Names for Blacks in the U.S.". http:// could include recent police brutality as well as mass names.mongabay.com/data/black.html incarceration. Both the extreme overrepresentation of Dawson, Michael C. Black Visions: The Roots of Contemporary African- American Political Ideologies. Chicago: University of Chicago, 2001. blacks in the criminal justice system, and the “Black "Support the Movement for Black Lives!" Black Lives Matter, accessed April Lives Matter” Movement in response to recent police 22, 2015, http://blacklivesmatter.com/. brutality, exemplify the way that blacks still face racial Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York: The New Press, 2010. distinction. These two phenomena demonstrate an existing racial structure, which arguably contributes Works Cited heavily to the monolithic nature of the black political 1. Michael C. Dawson, Behind the Mule: Race and Class in African- group. With distinctly racially biased events still American Politics (Princeton, NJ: Princeton UP, 1994), Print. occurring, it makes sense that African Americans 2. Michael C. Dawson, Behind the Mule, p 4. 3. Michael C. Dawson. Behind the Mule, p. 7. would continue to stick together politically on terms 4. "Blacks See Growing Values Gap Between Poor and Middle Class," of race. Pew Research Centers Social Demographic Trends Project RSS (Pew Research Center, 13 Nov. 2007). By better understanding the black community, 5. "Blacks See Growing Values Gap,” Pew Research Center. its history, and its politics, we can know more about 6. Theresa R. Kennedy. "Does Race Trump Socioeconomic Status in Shaping the African American Vote?" Junior Paper (Unpublished) how such a politically bonded group exists. This can (Princeton University, 2013), Print. inform political leaders, activists, and group members 7. Theresa R. Kennedy, “Does Race Trump Socioeconomic Status,” p. 9. about how to further engage black politics. Race is still 8. Ibid, pp. 15-16. 9. Ibid, p. 16. pertinent in American politics, as much, if not more so 10. Dawson, Michael C. Behind the Mule. 1994. P 4. than economic ideology, and this is something worth 11. "Ecological Inference." Ecological Inference. N.p., n.d. Web. 07 May 2013. paying attention to. Ultimately, we cannot truly know 12. Ibid. or understand American politics without furthering 13. Ibid.; Michael C. Dawson, Behind the Mule, p. 4. 14. Ibid, p. 4. our knowledge of black politics. 15. Claudine Gay, "Putting Race in Context: Identifying the Environmental Determinants of Black Racial Attitudes." (American Political Science Review 98.04, 2004). Bibliography 16. Michael C. Dawson, Behind the Mule, p 4. Dawson, Michael C. Behind the Mule: Race and Class in African-American 17. Ibid, p. 10. Politics. Princeton, NJ: Princeton UP, 1994. Print. 18. Ibid. Jackson, James S. National Black Election Panel Study, 1984 and 1988. 19. Ibid, p. 11. 20. Ibid, pp. 9-10. ICPSR09954-v1. Ann Arbor, MI: Inter-university Consortium for 21. Ibid, p. 11. 104 the salience of race versus socioeconomic status among african american voters

22. Ibid. 75. Claudine Gay, Putting Race in Context, p. 549. 23. Ibid, p. 35. 76. Michael C. Dawson, Black Visions, pp. xi-xii. 24. Ibid, p. 36. 77. Ibid. 25. Ibid. 26. Ibid. 27. Ibid, p. 10. 28. Michael C. Dawson, Black Visions: The Roots of Contemporary African-American Political Ideologies (Chicago: University of Chicago, 2001), p. xii. 29. "Blacks See Growing Values Gap,” Pew Research Center. 30. Ibid. 31. Ibid. 32. Ibid. 33. Ibid. 34. Lynette Clemetson, "Younger Blacks Tell Democrats to Take Notice," The New York Times, (The New York Times Company: 8 Aug. 2003), http://www.nytimes.com/2003/08/08/national/08VOTE.htmlhttp:// www.nytimes.com/2003/08/08/national/08VOTE.htm 35. Ibid. 36. "Civil Rights Movement," History.com, (A&E Television Networks: 1991), http://www.history.com/topics/black-history/civil-rights- Movement 37. Joann Wypijewski, "Black and Bruised," The New York Times (The New York Times Company: 1 Feb. 2004), http://www.nytimes. com/2004/02/01/magazine/01VOTERS.html, p. 5. 38. Ibid, p. 7. 39. Ibid, p. 8. 40. Claudine Gay and Jennifer L. Hochschild, “Is Racial Linked Fate Unique? Comparing Race, Ethnicity, Class, Gender, and Religion,” (2010), APSA 2010 Annual Meeting Paper, http://ssrn.com/ab- stract=1644497, pp. 5-6. 41. Ibid, p. 6. 42. Ibid, pp. 6-7. 43. Ibid, p. 16. 44. Ibid, p. 18. 45. Ibid. 46. Claudine Gay, "Putting Race in Context,” p. 547. 47. Ibid, p. 550. 48. Ibid. 49. Theresa R. Kennedy, "Does Race Trump Socioeconomic Status,” p. 9. 50. Ibid. 51. Ibid. 52. "Stanford Map of 2008 Precinct Winners," ArcGIS (ESRI: 2011), http://www.arcgis.com/home/item.html?id=87bb88214d2944bbbf1d 9d9255c812ca. 53. Theresa R. Kennedy, "Does Race Trump Socioeconomic Status,” pp. 15-16. 54. “Ecological Inference,” 2013. 55. James S. Jackson. National Black Election Panel Study. 1988. 56. Ibid. 57. Claudine Gay, Putting Race in Context, p. 547. 58. James S. Jackson. National Black Election Panel Study. 1988. 59. "Top 20 'Whitest' and 'Blackest' Names," ABC News (ABC News Net- work: 2006), http://abcnews.go.com/2020/story?id=2470131. 60. "Most Common Last Names for Whites in the U.S.," (N.p., 2000), http://Mongabay.com. 61. "Most Common Last Names for Blacks in the U.S.," (N.p., 2000), http://Mongabay.com. 62. James S. Jackson. National Black Election Panel Study. 1988. 63. Claudine Gay, Putting Race in Context. p. 549. 64. James S. Jackson. National Black Election Panel Study. 1988. 65. Michael C. Dawson, Behind the Mule, p. 4. 66. Claudine Gay, Putting Race in Context. p. 549. 67. Ibid, p. 550. 68. Ibid. 69. Ibid. 70. Ibid, p. 549. 71. Ibid. 72. Ibid, p. 558. 73. Michael C. Dawson, “Behind the Mule,” p. 4. 74. Lynette Clemetson, "Younger Blacks Tell Democrats," 2003. 105